Abstract : The prediction of social media information propagation is a problem that has attracted a lot of interest over the recent years, especially because of the application of such predictions for effective marketing campaigns. Existing approaches have shown that the information cascades in social media are small and have a large width. We validate these results for Tree-Shaped Tweet Cascades created by the ReTweet action. The main contribution of our work is a methodology for predicting the information diffusion that will occur given a user’s tweet. We base our prediction on the linguistic features of the tweet as well as the user profile that created the initial tweet. Our results show that we can predict the Tweet-Pattern with good accuracy. Moreover, we show that influential networks within the Twitter graph tend to use different Tweet-Patterns.
Lazaros Iliadis; Ilias Maglogiannis; Harris Papadopoulos. 10th IFIP International Conference on Artificial Intelligence Applications and Innovations (AIAI), Sep 2014, Rhodes, Greece. Springer, IFIP Advances in Information and Communication Technology, AICT-436, pp.79-89, 2014, Artificial Intelligence Applications and Innovations. 〈10.1007/978-3-662-44654-6_8〉
https://hal.inria.fr/hal-01391295
Contributeur : Hal Ifip
<>
Soumis le : jeudi 3 novembre 2016 - 10:51:35
Dernière modification le : mardi 26 décembre 2017 - 16:38:01
Document(s) archivé(s) le : samedi 4 février 2017 - 13:23:51
Eleanna Kafeza, Andreas Kanavos, Christos Makris, Pantelis Vikatos. Predicting Information Diffusion Patterns in Twitter. Lazaros Iliadis; Ilias Maglogiannis; Harris Papadopoulos. 10th IFIP International Conference on Artificial Intelligence Applications and Innovations (AIAI), Sep 2014, Rhodes, Greece. Springer, IFIP Advances in Information and Communication Technology, AICT-436, pp.79-89, 2014, Artificial Intelligence Applications and Innovations. 〈10.1007/978-3-662-44654-6_8〉. 〈hal-01391295〉
